9524050 Nepstad Modeling the effects of seasonal and interannual drought on the flow of water, carbon and nitrogen through Amazonian forests. Do trees of the Amazon forest experience significant drought stress? It seems that the answer to such a basic question should already be known, but, as is the case for many subjects in tropical ecology, our ignorance is vast. Drought stress in Amazonian forests is important because it determines how much water passes through the trees and enters the atmosphere. Knowing how much water the trees use will help us determine how cutting the forests and creating cattle pastures will affect flows of water in rivers and transfer of water vapor to the atmosphere. Water use is also important because it affects how well the trees grow and how much heat-trapping carbon dioxide the trees remove from the atmosphere. Amazonian droughts could become more frequent in the future if deforestation leads to a reduction in rainfall, as some climate models predict. If trees utilize water stored only in the top one or two meters of soil, as most ecologists and hydrologists currently assume, it is easy to calculate that water stress must develop during the 3-5 month dry periods that characterize most of Amazonia. However, most of Amazonia is covered by very deep soils that can store large quantities of water to several meters depth, and so the soils may protect the forest against dry seasons and against unusually dry years. If this is a general phenomenon, then Amazonian forests may experience significant drought stress only if a long-term decrease in rainfall occurs as the result of a general change in climate. In the proposed research, the hypothesis that Amazonian forests experience drought only during years of very low rainfall will be tested through field measurements of forests at three places where drought is more or less severe. The relationship between drought and the activity of microbes in the soil will also be examined at these sites. In order to integr ate the effects of drought on plants, microbes, and flows of water, and to make predictions about other locations in Amazonia and the tropics generally, these field measurements will be used to adapt a computer model to Amazon ecosystems. Through comparison of field measurements with model predictions, the effect of rooting depth, soil properties, and plant type on estimates of tropical forest water use and forest growth will be determined, laying the basis for an improved generation of climate and ecosystem models for the tropics. ??
